MiniDeepFont/Tests.py at master · amnonpaz/MiniDeepFont · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
from keras.preprocessing.image import load_img
from keras.preprocessing.image import img_to_array
from keras.preprocessing.image import ImageDataGenerator
import matplotlib.pyplot as plt
import cv2
import numpy as np
import h5py
import Preprocess
from math import floor
from CustomAugmentations import CustomAugmentations
import Database

def plot_normalized(hdf5_filename: str, sample_indices):
    rgb_colors = ['r', 'b', 'g']

    with h5py.File(hdf5_filename, 'r') as images_db:
        for sample_index in sample_indices:
            imgs_keys = list(images_db['data'].keys())
            key = imgs_keys[sample_index]
            img = images_db['data'][key][:]

            plt.figure('Image before/after normaization')
            plt.subplot()

            plt.subplot(2, 2, 1)
            plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB)), plt.xticks([]), plt.yticks([])
            plt.title('Original image')

            plt.subplot(2, 2, 2)
            for i in range(3):
                histogram, bin_edges = np.histogram(img[:, :, i], bins=256, range=(0, 256))
                plt.plot(bin_edges[0:-1], histogram, color=rgb_colors[i])

            img = Preprocess.normalize_image(img)
            plt.subplot(2, 2, 3)
            plt.imshow(cv2.cvtColor(img, cv2.COLOR_BGR2RGB)), plt.xticks([]), plt.yticks([])
            plt.title('Normalized image')

            plt.subplot(2, 2, 4)
            for i in range(3):
                histogram, bin_edges = np.histogram(img[:, :, i], bins=512, range=(-5, 5))
                plt.plot(bin_edges[0:-1], histogram, color=rgb_colors[i])

            plt.show()


def input_test(hdf5_filename: str, sample_index: int, shape: tuple, affine_crop=False, do_norm=True):

    with h5py.File(hdf5_filename, 'r') as images_db:
        imgs_keys = list(images_db['data'].keys())
        key = imgs_keys[sample_index]
        img = images_db['data'][key][:]

        '''
        for attr in images_db['data'][key].attrs:
            print('Available attribure {A}: {V}'.format(A=attr, V=images_db['data'][key].attrs[attr]))
        '''

        bboxes = images_db['data'][key].attrs['charBB']
        fonts = images_db['data'][key].attrs['font']
        letters = ''.join([word.decode('utf-8') for word in images_db['data'][key].attrs['txt']])
        print('Letters: {L}'.format(L=letters))

        # If we normalize, we want to see the difference between
        # the original image and the normalized image
        if do_norm:
            img = Preprocess.normalize_image(img)

        plt.figure('Fonts')
        plt.subplot()
        fonts_rows = 3
        fonts_to_display = min(bboxes.shape[-1], fonts_rows*3)
        for bb_idx in range(fonts_to_display):
            bb = bboxes[:, :, bb_idx]
            character = Preprocess.crop_character(img, bb, shape, affine_crop)

            plt.subplot(fonts_rows, floor(fonts_to_display/fonts_rows), bb_idx + 1)
            plt.imshow(character if not do_norm else cv2.cvtColor(character, cv2.COLOR_YUV2RGB))
            plt.title(letters[bb_idx] + ': ' + fonts[bb_idx].decode("utf-8")), plt.xticks([]), plt.yticks([])

        plt.show()


def plot_augmentations(filename:str, idx: int):
    train_x, train_y, _, _ = Database.load(filename)
    sample = np.expand_dims(train_x[idx], 0)
    datagen = CustomAugmentations()
    it = datagen.flow(sample, batch_size=1)
    for i in range(9):
        plt.subplot(330 + 1 + i)
        batch = it.next()
        image = batch[0]#.astype('uint8')
        plt.imshow(image)
    plt.show()


if __name__ == '__main__':
    input_filename = 'datasets/train/SynthText.h5'
    validation_filename = 'datasets/validation/SynthTextValidation.h5'
    font_train_db = 'datasets/train/ExtractedFonts.h5'
    font_validation_db = 'datasets/validation/ExtractedFontsValidation.h5'

    shape = (28, 28)
    input_test(input_filename , 450, shape, do_norm=True)
    #plot_normalized(input_filename , [45])